抄録
This study attempted to examine five types of questions in the English reading proficiency tests based on effectiveness of test-wiseness. Reading comprehension sections of the pre-2nd and 3rd grade of the STEP test (EIKEN) were used for the study. Based on question and answer relationships, questions were categorized into five types. These types were 'copy,' 'synonym,' 'paraphrase,' 'context,' and 'in-between.' The participants were 160 first-year and 69 third-year senior high school students. 'Written Recall' and 'Retrospective Interview' were adopted to gather protocols of test-takers. The data were classified and analyzed according to the list of test-taking strategies, in which strategies mainly dependent on language use strategies were marked 'Type A' and strategies mainly relying on test-wiseness strategies were named 'Type B.' The results suggest the participants decided how they would read the passage and answer the questions based on the test format. In other words, regardless of types of questions, test format may determine the way the participants deal with the test before they begin to answer questions. The results also imply that the questions which meet the following three conditions may have high validity. They are 1)a high rate of obtaining correct answers by utilizing Type A strategies, which means the text is easy enough for the examinees to read and understand, 2)no clues, except the contents of the passage, to reach correct answers,which makes Type B strategies less useful in acquiring correct answers, and 3)a less than 25% success rate of Type B strategies, which expresses that Type B strategies are not more effective than random guessing.
